1. Field of the Invention
The present invention relates to an image processing method of extracting a characteristic amount from a photographed image obtained, for example, by radiation (X-ray or the like) photography, and subjecting the photographed image to a gray scale conversion processing based on the characteristic amount.
2. Related Background Art
In recent years, with a progress in a digital technique, for example, a photographed image obtained by X-ray photography is digitized, and the digital image is subjected to an image processing and displayed in a monitor or outputted on a film for X-ray diagnosis.
Examples of the image processing include a gray scale conversion processing for converting a density of the photographed image to a density value at which the image is easily observed in a monitor image, film or another output device.
In the gray scale conversion processing, for example, when the photographed image obtained by photographing a chest with X-ray is outputted onto the film for X-ray diagnosis, first a histogram (pixel value histogram) of all pixels constituting the photographed image is prepared and analyzed. Moreover, a pixel value of a given partial point of the histogram (i.e., a pixel value obtained by subtracting a difference of 5% between a maximum pixel value and a minimum pixel value from the maximum pixel value in a pixel value distribution in the pixel value histogram, or the like) is extracted as the characteristic amount of the photographed image. The pixel value of the photographed image is converted (gray scale conversion) in such a manner that the extracted pixel value (characteristic amount) corresponds to a given density value (density of about 1.9) on the film.
However, in the aforementioned conventional image processing method, when the characteristic amount is extracted by histogram analysis, the pixel value of the given partial point (higher 5% point or the like) of the histogram is extracted as the characteristic amount. However, the pixel value of the given partial point is often a pixel value in an area of a portion different with a photographic subject.
Concretely, for example, a characteristic amount extracted from a predetermined portion in a lung is most appropriate in the photographed image of the chest (the extracted characteristic amount corresponds to the pixel value of the predetermined portion in the lung). However, the characteristic amount is extracted from a higher portion in the lung (corresponds to the pixel value of the higher portion in the lung) or the characteristic amount is extracted from a soft tissue portion (corresponds to the pixel value of the soft tissue portion) depending upon the photographic subject. Moreover, when a plaster cast is added to the photographic subject, the characteristic amount is extracted from another unrelated portion (corresponds to the pixel value of another unrelated portion) in some case.
As described above, in the conventional art, even when the pixel value of the given partial point of the histogram is extracted as the characteristic amount, the extraction area of the characteristic amount differs with each photographic subject. Therefore, even when the gray scale conversion processing is performed based on the characteristic amount extracted from the photographed image, a dispersion of density distribution is generated in each processed image. If the image subjected to the gray scale conversion and having the dispersion in the density distribution is used to perform diagnosis, a diagnosis mistake is possibly caused. This raises a very large problem.